processor(prompt, images=image) to `processor(images=image, text=pr…

…ompt)` (#2076)

* `prompt, images=image` to `images=image, text=prompt`

* special case of model being str in BCO

trl/trainer/bco_trainer.py

@@ -336,7 +336,7 @@ def __init__(
         if type(args) is TrainingArguments:
             raise ValueError("Please use `BCOConfig` instead `TrainingArguments`.")
 
-        if ref_model is model:
+        if not isinstance(model, str) and ref_model is model:
             raise ValueError(
                 "`model` and `ref_model` cannot be the same object. If you want `ref_model` to be the "
                 "same as `model`, you must mass a copy of it, or `None` if you use peft."

trl/trainer/dpo_trainer.py

@@ -123,7 +123,7 @@ def _process_prompt(
         )
         prompt_tokens = []
         for prompt, image in zip(prompts, images):
-            tokens = processor(prompt, images=image, **processor_kwargs)
+            tokens = processor(images=image, text=prompt, **processor_kwargs)
             tokens = {k: v[0] for k, v in tokens.items()}
             if not isinstance(tokens["input_ids"], list):
                 tokens["input_ids"] = tokens["input_ids"].tolist()
@@ -302,7 +302,7 @@ def tokenize(text, images=None):
                 if "add_special_tokens" in inspect.signature(processor).parameters
                 else {}
             )
-            tokenized = processor(text, images=images, **processor_kwargs)
+            tokenized = processor(images=images, text=text, **processor_kwargs)
             tokenized = {k: v[0] for k, v in tokenized.items()}
             if not isinstance(tokenized["input_ids"], list):
                 tokenized["input_ids"] = tokenized["input_ids"].tolist()